Risk Assessment of Hydroclimatic Variability on Groundwater Levels in the Manjara Basin Aquifer in India Using Archimedean Copulas

Abstract

In this paper, a bivariate-copula-based methodology is presented to assess the risk associated with hydroclimatic variability on groundwater levels in an unconfined aquifer at the Manjara basin in India. Rank correlation analysis is used to identify the association between the El Niño–Southern Oscillation (ENSO) index, precipitation, and groundwater levels. It is found that the dependencies among the hydroclimatic variable pairs are statistically significant and the dependence structure can be modeled by using bivariate Archimedean copulas. The groundwater level or depth-to-groundwater table (DGWT) in the study region is found to be responsive toward interannual precipitation variations that are influenced by the ENSO phenomenon. For probabilistic representation of hydroclimate variables, various probability distributions are evaluated and it is found that the precipitation and DGWT are best fitted using lognormal and Weibull distributions, respectively, whereas the ENSO index is best fitted using nonparametric-based normal kernel density function. For modeling joint dependence structure of hydroclimatic variable pairs (precipitation-DGWT, ENSO index-precipitation, and ENSO index-DGWT), appropriate Archimedean copulas (viz, Ali-Mikhail-Haq, Clayton, Gumbel-Hougaard, and Frank families) are evaluated. On performing standard statistical tests, it is found that the Frank copula is best representing the joint dependence structure for all three variable pairs. Then the Frank copula-based joint distributions are used to derive conditional distributions and to perform risk analysis of groundwater levels. The study suggest that the copula-based methodology can be used effectively for modeling dependence structure of hydroclimatic variables and for risk assessment of groundwater levels under changes in hydroclimatic conditions.